CN110096762A - A kind of prediction of lathe rigging error and control method - Google Patents

Abstract

The invention discloses a kind of prediction of lathe rigging error and control methods, comprising the following steps: (1) based on Hertzian contact theory, establishes rolling guide error-workbench Error Propagation Model based on compatibility of deformation, and linearized；(2) it is directed to precise horizontal machining center, establishes finite element model by ABAQUS simulation software, extracts the gravity deformation result of each assembly step；(3) based on the assembly Error Propagation Model based on differential vector method, the complete machine error model for considering assembly deflections and guiding error is established；(4) it is directed to established complete machine error model, proposes corresponding rigging error adjustment, control strategy.The present invention has comprehensively considered influence of the rolling guide geometric error to error propagation, is more in line with objective installation situation；The changeability for considering gravity deformation, gradually analyzes deformation result, more accurately accomplishes gradually to predict and control end error.

Description

A Machine Tool Assembly Error Prediction and Control Method

technical field

The invention relates to the field of assembling and testing of numerically controlled machine tools, in particular to a method for predicting and controlling machine tool assembling errors taking into account rolling guide rail errors and gravity deformation of structural parts.

Background technique

High-speed, high-precision, and high-reliability precision machining centers have become the development direction of modern equipment manufacturing [1]. Many countries regard the development of precision CNC machine tools as the primary task of developing high-end manufacturing. As an important CNC machine tool, the precision horizontal machining center has a series of advantages such as high degree of automation and high processing efficiency, and is widely used in aerospace, aviation, precision mold processing and other fields.

There is no clear and effective guidance method for guide rail installation and adjustment, which takes up a lot of time in the assembly process. Technicians often ignore the effect of guide rail error on the error transmission of sliders and moving parts under the effect of geometric errors such as straightness on the rolling joint surface. At the same time, the gravity deformation of parts during the assembly process has a great influence on the assembly error. At present, the anti-deformation strategy is often used to compensate for the assembly work of machine tools, but the control amount of anti-deformation is not very accurate, and the transfer effect of the guide rail error is not considered. The functional relationship between the deformation deviation and the homogenization of the guide rail and the final assembly error cannot be established.

Therefore, the current traditional assembly process mainly relies on the experience of workers, lacks scientific theoretical analysis and guidance specifications, makes assembly reliability difficult to guarantee, and assembly efficiency is low.

[1] Liu Jia. Research on Modeling and Control Technology of Assembly Failure Rate of CNC Machine Tool [D]. Chongqing: Chongqing University, Master Thesis, 2012.

Contents of the invention

The object of the present invention is to overcome the deficiencies in the prior art, and provide a machine tool assembly error prediction and control method that considers the error of the rolling guide rail and the gravity deformation of the structural parts. Through the research on the assembly error transmission mechanism of the precision horizontal machining center, comprehensively considering the influence of the guide rail error and the gravity deformation, the assembly error transmission model is established, and the assembly error control adjustment method is proposed to provide a theory for the assembly error prediction control of the precision horizontal machining center According to the scientific guidance of machine tool assembly personnel, the assembly accuracy and assembly speed of my country's precision machine tools have been greatly improved.

The purpose of the present invention is achieved through the following technical solutions:

A machine tool assembly error prediction and control method, comprising the following steps:

(1) Based on the Hertz contact theory, a rolling guide error-table error transfer model based on deformation coordination is established and linearized;

(2) For the precision horizontal machining center, use the ABAQUS simulation software to establish a finite element model, and extract the gravity deformation results of each component under each assembly step;

(3) Based on the assembly error model based on the differential vector method, an assembly error transfer model of the whole machine is established considering the gravity deformation of each component and the error of the guide rail;

(4) According to the established assembly error transfer model of the whole machine, put forward the corresponding assembly error adjustment and control strategy.

Further, in step (1), the establishment of a rolling guide error-worktable error transfer model based on deformation coordination specifically includes the following steps:

(101) Establish a coordinate system and define errors at the center of gravity of the moving parts, on the table and at the center of each slider;

(102) solving roller deformation and contact force;

(103) Establishing an error transfer model according to the force balance;

(104) Linearize the error transfer model.

Further, step (2) specifically includes the following steps:

(201) setting the assembly step and selecting the spatial position;

(202) Extract the gravity deformation of each assembly step part;

(203) Calculate straightness deviation and angle deviation caused by gravity deformation.

Further, step (3) specifically includes the following steps:

(301) Define the assembly joint surface between the sub-assemblies as the key product feature in the assembly error transfer model; define the geometric deviation state of each key product feature;

(302) Taking an assembly composed of three parts as an example, deduce an assembly error transfer model;

(303) Establish an assembly error transmission model of the horizontal machining center.

Further, the key product features in step (301) include: the deviation of the joint surface of the bed column relative to the reference coordinate system, the perpendicularity deviation of the X-axis guide rail installation plane of the column relative to the joint surface of the bed column, and the slider of the X-axis guide rail. Surface deviation, the parallelism deviation of the installation plane of the Y-axis guide rail of the slide plate relative to the joint surface of the slide plate slider, the deviation of the slider surface of the Y-axis guide rail, the parallel deviation of the spindle end relative to the joint surface of the headstock slider, and the bed The deviation of the Z-axis guide rail installation plane relative to the reference coordinate system, the deviation of the slider surface of the Z-axis guide rail, and the parallelism deviation of the upper surface of the worktable relative to the joint surface of the worktable slider.

Further, step (4) specifically includes the following steps:

(401) Measure the assembled assembly to obtain its deviation state; under the deviation state of the existing assembly, assuming that there is no error in the unassembled parts, predict the deviation state of the end of the whole machine, and obtain only the assembled Prediction results of machine assembly errors for component errors;

(402) Measure the unassembled parts to obtain their geometric deviation. Under the deviation state of the existing assembly, use the deviation measurement value of the unassembled parts instead of the actual error of the unassembled parts to predict, and all subassemblies are considered Prediction results of assembly errors of the whole machine based on body errors;

(403) If each item of the prediction result of the assembly error of the whole machine that only considers the error of the assembled parts has exceeded the required deviation state, adjust the assembled parts;

(404) If the predicted result of the assembly error of the whole machine obtained by considering only the error of the assembled parts does not exceed the required deviation state, continue to judge whether the predicted result of the assembly error of the whole machine under the condition of considering all sub-assembly errors exceeds the required deviation state , if it has been exceeded, the unassembled parts should be adjusted; until step (403) and step (404) do not exceed the deviation requirement of the target feature, then proceed to the next step of assembly until the assembly is completed.

Compared with the prior art, the beneficial effects brought by the technical solution of the present invention are: the present invention comprehensively considers the influence of the geometric error of the rolling guide rail on the assembly error of the complete machine, and the established assembly error prediction model of the complete machine is consistent with the actual assembly error. The transfer process is more appropriate and more in line with the objective assembly situation; considering the variability of gravity deformation, the deformation results of each component under different assembly steps are extracted in step (2), and the assembly error transfer model established in step (3) Taking into account the gravity deformation under different assembly steps, it can be predicted step by step; combined with the precision machine tool assembly strategy proposed in step (4), the error of each component can be gradually controlled during the assembly process, reducing repeated scraping work, and can accurately and efficiently guide machine tool assembly Work.

Description of drawings

Figure 1 is a schematic diagram of the coordinate system of the workbench and the guide rail slider;

Figure 2 is a schematic diagram of the assembly sequence of the machine tool;

Fig. 3 is the schematic diagram of the position of each component in the simulation;

Fig. 4 is the Y-direction displacement nephogram of the column in each assembly step;

Fig. 5 is the Z-direction displacement nephogram of the bed in each assembly step;

Figure 6(a) is the Z-direction straightness error curve of the guide rail on the X-axis that changes with the assembly step;

Figure 6(b) is the Y-direction straightness error curve of the guide rail on the X-axis that changes with the assembly step;

Figure 7(a) is the Z-direction straightness error curve of the guide rail under the X-axis that changes with the assembly steps;

Figure 7(b) is the Y-direction straightness error curve of the guide rail under the X-axis that changes with the assembly steps;

Figure 8 is the error value of the rotation angle around X of the guide rails of each axis that changes with the assembly step;

Figure 9 is a schematic diagram of the coordinate system and geometric deviation of the key features of the bed and the Z-axis guide rail;

Figure 10 is a schematic diagram of the coordinate system and geometric deviation of the key features of the column and the X-axis guide rail;

Figure 11 is a schematic diagram of the coordinate system and geometric deviation of the key features of the slide plate and the Y-axis guide rail;

Figure 12 is a schematic diagram of the coordinate system and geometric deviation of the key features of the headstock and workbench;

Fig. 13 is the transmission process of geometric deviation and deformation deviation of three-part assembly;

Figure 14 is a flow chart of the machine tool assembly error control strategy.

Detailed ways

The present invention will be described in further detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.

1 Rolling guide error-worktable error transfer model establishment

1.1 Hertz Contact Theory

According to the Hertz contact theory, the relationship between contact deformation and contact pressure is:

In engineering applications, because the force of the roller is very complicated, its empirical formula is mainly used (reference: Zhupanska O I. Contact problem for elastic spheres: Applicability of the Hertz theory to non-small contact areas[J].International Journal of Engineering Science, 2011, 49(7):576-588.):

Among them, E ₁ , E ₂ are the elastic modulus of the roller and the raceway surface; υ ₁ , υ ₂ are the Poisson's ratio of the roller and the raceway surface; R is the radius of the roller; l is the length of the roller.

Rollers and raceways are generally made of steel, and their elastic modulus E and Poisson's ratio υ are the same, then the contact force Q is a function of contact deformation δ _n :

in,

For the rolling element, there are two contact points, and the deformation occurs at the two contact points at the same time, which belongs to bilateral deformation. Therefore, according to the empirical formula of the Hertz contact theory, the contact force of the rolling element ki can be obtained:

Among them, F _ki is the contact force of the rolling elements in the k raceway on the i-th section, and its direction depends on the contact angle. Δd _ki is the total deformation of the roller, E and υ are the elastic modulus and Poisson's ratio of the ball and the raceway material, respectively.

1.2 Error transfer model between slider and moving parts

1.2.1 Establishment of coordinate system and definition of each error

For the structure of double guide rails and four sliders commonly used in machine tools at present, firstly, a coordinate system is established at the center of gravity of the moving parts, the table top and the center of each slider, as shown in Figure 1. The center of gravity of the worktable is relative to the installation plane, that is, the distance from the upper surface of the slider is H ₁ , and the distance from the center of gravity of the table top coordinate system is H ₂ . The span of the guide rail is L _v , and the distance between two sliders on the same guide rail is L _h . Assume that the coordinates of the origin of the coordinate system of the four sliders in the coordinate system of the moving part are [x _r,j ,y _r,j ,z _r,j ] (j=1,2,3,4).

Since the guideway error will cause the position and attitude errors of the moving parts, it is necessary to assume separately:

In the slider coordinate system, the guide rail error is:

Δ _g,j =[δ _x,gj ,δ _y,gj ,θ _x,gj ,θ _y,gj ,θ _z,gj ] ^T (7)

In the coordinate system of the moving part, the error of the center of the moving part is:

Δ _t = [δ _x,t ,δ _y,t ,θ _x,t ,θ _y,t ,θ _z,t ] ^T (8)

1.2.2 Roller deformation and contact force solution

Without considering the deformation of the moving parts, in the coordinate system of the moving parts, the contact point A _r,jki =[x _r,jki ,y of the slider raceway k and the roller on any section i in the slider j _r,jki ,z _r,jki ]T, will move to A _rj ′=[x _r,jki ′,y _r,jki ′,z _r , _jki ′]T due to the influence of the error, namely:

Similarly, in the slider coordinate system, the error of the guide rail will cause the contact point A _g of the guide rail raceway k and the roller on any section i of the slider j = [x _g,jki ,y _g,jki ,z _g,jki ]T will move to A _g ′=[x _g,jki ′,y _g,jki ′,z _g,jki ′]T, then:

Unifying it into the coordinate system of moving parts, the distance change between A _g and A _rj ′ can be obtained as:

Where Δd _pre is the preload of the rolling body, x _{r, j} and y _{r, j} are the X and Y coordinates of the origin of the coordinate system of the j slider in the coordinate system of the moving part.

According to formulas (5)-(6), the contact force of the roller can be obtained as:

1.2.3 Force balance analysis to establish error transfer model

In the coordinate system of the moving part, the resultant contact force and resultant moment of all rollers can be obtained, assuming that the center of the moving part is subjected to the external force F _x,t ,F _y,t and the external force distance M _x,t ,M _y,t ,M _{z ,t} , according to the force balance of the system, the following balance equation can be established:

According to the formula (13), it is known that the error Δ _g,j of the position of the four sliders =[δ _x,gj ,δ _y,gj ,θ _x,gj ,θ _y,gj ,θ _z,gj ] ^T , Simultaneous formulas (9)-(13) can obtain the five-dimensional error Δt=[δ _x,t ,δ _y,t ,θ _x,t ,θ _y,t ,θ _z,t ] ^T of the moving parts.

The following formula (14) is used to represent the relationship between the error of the moving parts and the respective errors of the four sliders, namely:

_Δt = G(Δg _,1, Δg _,2, Δg _,3 ,Δg _,4 ) (14)

Δ _g,j =[δ _x,gj ,δ _y,gj ,θ _x,gj ,θ _y,gj ,θ _z,gj ] ^T is the error of slider j, Δ _t =[δ _x,t ,δ _{y ,t} ,θ _x,t ,θ _y,t ,θ _z,t ] ^T is the error of the moving parts.

1.2.4 Error transfer model linearization

The above model is a nonlinear model, although it can more accurately describe the transfer relationship between the error of the four sliders and the error of the moving parts. However, because the nonlinear relationship is difficult to deal with in the analysis, it is not conducive to the prediction of the assembly error of the whole machine, so the method of multiple linear regression is used for linearization. Determine the necessary variables by performing significance analysis, and then use Matlab to obtain the linear regression matrix $, that is, the relationship between Δ _t and Δ _g,1 , Δ _g,2 , Δ _g,3 , Δ _g,4 satisfies the following relationship:

Therefore, the error transfer linearization model between the moving part and the slider is established, and the error Δ _g,j of the position of the four sliders is known = [δ _x,gj ,δ _y,gj ,θ _x,gj ,θ _{y ,gj} ,θ _z,gj ] ^T , the five-dimensional error Δ _t = [δ _x,t ,δ _y,t ,θ _x,t ,θ _y,t can be obtained by formula (15) ,θ _z,t ] ^T .

2. Finite element simulation of machine assembly deformation

During the machine tool assembly process, the gravity of each structural part will cause deformation or displacement of the joint surface, which will affect the assembly error of the machine tool. Since the assembly deformation cannot be obtained by means of measurement, the gravity deformation of the structural parts is extracted through the Abaqus simulation software below.

2.1 Assembly step setting and spatial location selection

Since each structural part is assembled step by step, the results of gravity deformation are different with the change of assembly steps, and the results of gravity deformation need to be obtained for each assembly step. The content of each assembly step is shown in Figure 2; at the same time, due to When the moving structural parts are in different spatial positions, they will affect the static deformation state of the guide rail to varying degrees. Three positions are selected for each guide rail axis (as shown in Figure 3, the three positions selected for each guide rail are respectively 0, 1 and 2), the simulation data of three positions can be obtained, and then the remaining positions can be calculated by polynomial fitting, and the deformation or displacement of any position can be obtained.

2.2 Deformation extraction of components in each assembly step

Taking the upper and lower guide rails of the column as an example, the deformation cloud diagram obtained in each assembly step is shown in Figure 4-5.

2.3 Calculate the straightness and angle deviation caused by gravity

Through the displacement change results of the finite element analysis, the straightness error of the guide rail and the angular error around the X axis caused by the deformation in the assembly step can be separated. For straightness error, simply subtract the lowest point on the same rail. For the rotation angle error around the X axis, the X-axis guide rail is the difference between the end point of the upper guide rail and the same end point of the lower guide rail, and then divided by the span of the guide rail, and the Y and Z-axis guide rails are the difference between the two ends of the same guide rail divided by the length of the guide rail.

Taking the column guide rail as an example, the straightness results after separation are shown in Figure 6(a) to Figure 7(b), and the rotation angles of the guide rails around the X axis are shown in Figure 8.

3. Establish the error model of the whole machine based on the differential vector method

3.1 Definition of geometric deviation state of each feature

Since the deviation of the whole assembly is mainly determined by the joint surface deviation between subassemblies, the assembly joint surface between subassemblies is usually defined as a key product feature in the assembly error transfer model. Table 1 shows all the key product features defined, and the specific positions of the features are shown in Figure 9-12.

Table 1 Symbols and meanings of the geometric deviations of each joint surface

In three-dimensional space, each feature has six degrees of freedom, that is, translational degrees of freedom along three coordinate axes and rotational degrees of freedom around three coordinate axes. Furthermore, the deviation state of feature k relative to its theoretical position and orientation can be represented by a 6×1 vector, namely:

[P _k Q _k ] ^T = [ΔX _k ,ΔY _k ,ΔZ _k ,Δθ _xk ,Δθ _yk ,Δθ _zk ] ^T (16)

Among them, k represents the feature number.

Considering the simultaneous existence of shape error and gravity deformation, for the deviation state of the feature, it is necessary to introduce i to represent the assembly step, namely:

Define when i=0 is the unassembled state, when i=k is the state after the kth part is assembled.

First, since the geometric deviation of the feature itself does not change with the assembly step, it is defined as:

Among them, δ _k is the differential translation vector, and ε _k is the differential rotation vector.

The deviation due to the deformation of the feature due to gravity varies with the assembly step and is written as:

in, is the differential translation vector, Differential rotation vector.

3.2 Taking an assembly composed of three parts as an example, deduce the general form of the assembly error transfer model

It can be known from the differential vector method that:

R is a 3x3 rotation matrix, D is a 3x3 translation matrix, and d is a distance vector in the corresponding direction. The upper and lower subscripts indicate the two coordinate systems in which the transformation takes place.

As shown in Figure 13:

In the first step of assembly, surface 1 has not yet been deformed, only geometric deviation:

In the second step of assembly, the deviation caused by gravity deformation is superimposed on the geometric deviation, and the deviation of surface 1 is:

And the deviation of surface 2 is:

Similarly, the deviation state of feature k after the i-step assembly can be extended as:

After the assembly of the whole machine is completed, that is, after the n-step assembly is completed, the deviation state of feature k is:

It should be noted that since the deformation deviation state is related to the assembly step, the intermediate state of feature k is different from the state after the completion of assembly, namely:

When all n assembly sub-bodies of the assembly are assembled, the status of all features changes to Then the deviation of feature n is:

3.3 Establishment of assembly error transfer model of precision horizontal machining center

Since the precision horizontal machining center is a closed-loop assembly composed of two open-loop assemblies, the deviation of the main shaft and the deviation of the worktable can be calculated separately to obtain the assembly error transfer model of the whole machine. The following mainly takes the deviation state of the main shaft as an example for derivation.

When the joint surface of the bed column has a geometric deviation, the deviation state is:

According to formula (25), when the column and the X-axis guide rail are assembled, the deviation state of the slider surface of the X-axis guide rail is not only affected by its own geometric deviation, but also by the joint surface of the bed column and the installation surface of the X-axis guide rail, so X The deviation state of the shaft guide slider is:

The subscript sj represents the slider j of the skateboard; δ _sj and ε _sj are the position error and rotation angle error vectors of the slider respectively.

According to the linear regression error model between the skateboard and the slider that drives it, the deviation state of the skateboard can be deduced as:

Among them, B _sj (j=0,1..4) is the matrix corresponding to slider j in the skateboard regression coefficient matrix.

Similarly, considering the influence of the slider error and the headstock slider error on the spindle, the deviation state of the spindle is:

Among them, B _bj represents the linear regression coefficient matrix of the headstock slider j, and δ _bj and ε _bj are the displacement error and rotation angle error vectors of the headstock slider respectively.

Similarly, the deviation state of the workbench is:

A _t,mj =B _tj W _5,tj ,A _t,nj =B _tj W _6,tj ,A _t,kj =B _tj W _tj,tj

Among them, B _tj represents the linear regression coefficient matrix of the worktable slider j, and δ _tj and ε _tj are the displacement error and rotation angle error vectors of the worktable slider respectively.

The precision horizontal machining center is a closed-loop assembly composed of two assembly open loops. Therefore, by rotating the deviation of the spindle and the worktable to the same coordinate system, the relative deviation between the worktable and the spindle can be obtained as:

4. Assembly error prediction and control strategy

Process-oriented assembly error control is a method to control the error of each assembly step to achieve the final error to meet the requirements. According to the established model, in the process of machine tool assembly, the deviation and error of the target feature can be predicted for each installation, and at the same time, the degree of influence of the assembled parts on the target feature can be analyzed, so the assembly can be as follows error control strategy.

Step 1: Measure the assembled assembly to obtain its deviation state [P _k , Q _k ] ^T ; under the deviation state [P _k , Q _k ] ^T of the existing assembly, it is assumed that the unassembled parts have no Error, predict the state of the end deviation, and obtain the prediction result of the assembly error of the whole machine that only considers the error of the assembled body

The second step: measure the unassembled parts to obtain their geometric deviation [δ _k+1 ,ε _k+1 ] ^T . Under the deviation state [P _k , Q _k ] ^T of the existing assembly, the deviation measurement value of the unassembled parts is used to replace the actual error of the unassembled parts for prediction, and the assembly error prediction of the whole machine is obtained considering the errors of all sub-assemblies result

Step 3: If only the error of the assembled parts is considered, the prediction result of the assembly error of the whole machine Items of have exceeded the required deviation status Then the assembled parts must be adjusted.

Step 4: If only the error of the assembled parts is considered, the prediction result of the end error of the whole machine not exceeded Then continue to judge the assembly error prediction results of the whole machine considering all sub-assembly errors Is it more than If exceeded, adjustments should be made to unassembled components. It can be analyzed by analyzing two prediction results ( and ), determine the source of each error and whether it is greatly affected by the assembled parts or the unassembled parts, and adjust the error source with a large influence until the third step and the fourth step do not exceed the target feature If the deviation requirements are met, proceed to the next step of assembly until the assembly is completed.

The process flow of the above assembly error control strategy is shown in Figure 14 below.

The present invention is not limited to the embodiments described above. The above description of the specific embodiments is intended to describe and illustrate the technical solution of the present invention, and the above specific embodiments are only illustrative and not restrictive. Without departing from the gist of the present invention and the scope of protection of the claims, those skilled in the art can also make many specific changes under the inspiration of the present invention, and these all belong to the protection scope of the present invention.

Claims (6)

1. A machine tool assembly error prediction and control method, is characterized in that, comprises the following steps: (1) Based on the Hertz contact theory, a rolling guide error-table error transfer model based on deformation coordination is established and linearized; (2) For the precision horizontal machining center, use the ABAQUS simulation software to establish a finite element model, and extract the gravity deformation results of each component under each assembly step; (3) Based on the assembly error model based on the differential vector method, an assembly error transfer model of the whole machine is established considering the gravity deformation of each component and the error of the guide rail; (4) According to the established assembly error transfer model of the whole machine, put forward the corresponding assembly error adjustment and control strategy. 2. a kind of machine tool assembly error prediction and control method according to claim 1, is characterized in that, in step (1), set up the rolling guide rail error-worktable error transfer model based on deformation coordination and specifically include the following steps: (101) Establish a coordinate system and define errors at the center of gravity of the moving parts, on the table and at the center of each slider; (102) solving roller deformation and contact force; (103) Establishing an error transfer model according to the force balance; (104) Linearize the error transfer model. 3. A kind of machine tool assembly error prediction and control method according to claim 1, is characterized in that, step (2) specifically comprises the following steps: (201) setting the assembly step and selecting the spatial position; (202) Extract the gravity deformation of each assembly step part; (203) Calculate straightness deviation and angle deviation caused by gravity deformation. 4. A kind of machine tool assembly error prediction and control method according to claim 1, is characterized in that, step (3) specifically comprises the following steps: (301) Define the assembly joint surface between the sub-assemblies as the key product feature in the assembly error transfer model; define the geometric deviation state of each key product feature; (302) Taking an assembly composed of three parts as an example, deduce an assembly error transfer model; (303) Establish an assembly error transmission model of the horizontal machining center. 5. A method for predicting and controlling machine tool assembly errors according to claim 4, wherein the key product features in the step (301) include: the deviation of the joint surface of the bed column with respect to the reference coordinate system, the X-axis guide rail installation of the column The deviation of the verticality of the plane relative to the joint surface of the bed column, the deviation of the slider surface of the X-axis guide rail, the parallelism deviation of the installation plane of the Y-axis guide rail of the slide plate relative to the joint surface of the slide plate slider, and the deviation of the slider surface of the Y-axis guide rail. Deviation, the deviation of the parallelism of the spindle end relative to the joint surface of the spindle box slider, the deviation of the installation plane of the Z-axis guide rail of the bed relative to the reference coordinate system, the deviation of the slider surface of the Z-axis guide rail and the upper surface of the table relative to the working surface Parallelism deviation of the joint surface of the table slider. 6. A kind of machine tool assembly error prediction and control method according to claim 1, is characterized in that, step (4) specifically comprises the following steps: (401) Measure the assembled assembly to obtain its deviation state; under the deviation state of the existing assembly, assuming that there is no error in the unassembled parts, predict the deviation state of the end of the whole machine, and obtain only the assembled Prediction results of machine assembly errors for component errors; (402) Measure the unassembled parts to obtain their geometric deviation. Under the deviation state of the existing assembly, use the deviation measurement value of the unassembled parts instead of the actual error of the unassembled parts to predict, and all subassemblies are considered Prediction results of assembly errors of the whole machine based on body errors; (403) If each item of the prediction result of the assembly error of the whole machine that only considers the error of the assembled parts has exceeded the required deviation state, adjust the assembled parts; (404) If the predicted result of the assembly error of the whole machine obtained by considering only the error of the assembled parts does not exceed the required deviation state, continue to judge whether the predicted result of the assembly error of the whole machine under the condition of considering all sub-assembly errors exceeds the required deviation state , if it has been exceeded, the unassembled parts should be adjusted; until step (403) and step (404) do not exceed the deviation requirement of the target feature, then proceed to the next step of assembly until the assembly is completed.